To ensure data is available in real-time customer profiles with less than one minute latency, which ingestion type is recommended?

Streaming ingestion is the recommended method for ensuring that data is available in real-time customer profiles with less than one minute latency. This type of ingestion allows for continuous data flow into the system, meaning as new data is generated or received, it can be processed and made readily available for use almost immediately.

The real-time capabilities of streaming ingestion are designed to handle continuous input efficiently, making it essential for applications requiring up-to-the-moment data availability, such as customer profile updates and personalization experiences. By utilizing this method, organizations can react swiftly to changes in customer interactions, preferences, and behaviors, ensuring they deliver a timely and relevant experience to users.

In contrast, the other ingestion types—batch, scheduled, and import ingestion—typically involve delays. Batch ingestion collects data over a period before processing it together, which inherently creates latency. Scheduled ingestion processes data at set intervals, also resulting in delays. Import ingestion might refer to transferring data from external sources but, like batch and scheduled ingestion, does not support the continuous flow of data necessary for real-time applications. Thus, streaming ingestion stands out as the effective choice for achieving low-latency data availability in customer profiles.